馬可夫鍊理論應用於龜山島海域波浪預報之研究

Application of Markov Chain Theory to Forecast Wind Wave over Gueishandao Water

李汴軍(Beng-Chun Lee) ； 吳立中(Li-Chung Wu) ； 林家豐(Chia-Feng Lin) ； 董東璟(Dong-Jiing Doong)

《海洋工程學刊》 7卷1期 (2007/06) Pp. 69-84

https://doi.org/10.6266/JCOE.2007.0701.04

摘要

本文應用馬可夫鍊移轉機率矩陣的方法，進行龜山鳥海域短時期波浪預報的研究，來輔助海上遊憩以及登島作業所需海況資訊之需求。爲能獲得符合該海域波浪特性之馬可夫鍊移轉機率矩陣，本文透過龜山鳥海域長期問的實測風浪資料進行迴歸分析，作爲馬可夫鍊波高移轉機率矩陣分類依據。預報結果顯示，在3天內波高預測之誤差平均值在30 cm以內。本研究進一步嘗試結合風速及波高的狀態，建立出更爲複雜的風速波高聯合馬可夫鍊矩陣，分析結果發現，此一風速波高聯合馬可夫鍊矩陣適用於現場風速爲1~6級時的波浪預報。從不同季節的波浪預報結果發現，波高預報準確性以冬季最高。颱風期間波浪預報的誤差較其它季節大，這是因爲颱風期問波浪資料序列的非定常性較強烈，導致馬可夫鍊理論預報的誤差會明顯偏大。內於颱風期間一律禁止遊客登島，颱風期間的預報誤差，並不實際作爲登島決策的依據。對於龜山島登島作業的決策，馬可夫鍊法有其實用的價值。

關鍵字

馬可夫鍊理論 ； 波浪預報

並列摘要

By adopting the transition probability matrix of Markov chain method, this paper attempts to predict the short-term wave height in order to fulfill the demand of visiting the island. The result show that the mean error of 3 days forecast of the prediction wave height is less than 30 cm. It also reveals that the joint probability of transition matrix of Markov chain, which based on the observed wind speed and wave data, is suitable for the wave forecast during the Beaufort scale 1~6 grade. From various waves forecast in different seasons it is known that the accuracy of the forecast wave height is relatively high in winter. There might be a great error for waves forecast during typhoon season, as its nonstationary characteristics, which might influence the accuracy of transition probability matrix of Markov chain. Nevertheless, as visitors are forbidden to enter the island during typhoon season, there is no need to consider the marginal error of typhoon season. In sum, this paper demonstrates the application of Markov chain theory and argues that it is useful to the wave forecast over Gueishandao water.

並列關鍵字

Markov chain theory ； wave forecast